This implements de/serialization for the types that aren't templated (SwerveDriveKinematics) in C++ or where there is no trivial way to go round-trip (Splines) for the messages.
To reduce the need for users to manually perform unit conversions, this allows Measure objects from wpiunits to be passed into most places in wpimath that currently expect doubles in terms of SI units like meters.
For example, users would need to know that unit conversion is required - and what the correct units are. Using units would be more difficult to write code for than just hardcoding a value or using Units.inchesToMeters.
Now, using units has no more developer overhead than using raw numbers.
ProfiledPIDController and ExponentialProfile use current, then goal.
This isn't a breaking change because this overload of calculate() is
new for 2024.
Made JNI modifications to expose the faster function, made the API use
the typesafe Matrix API, and synchronized the documentation with C++.
Sped up C++ LTV diff drive test from 20 ms to 15 ms.
Sped up C++ LTV unicycle test from 15 ms to 10 ms.
Both seem to work, but the SDA algorithm is specifically recommended for
solving DAREs as opposed to P-DAREs.
The QR decomposition was replaced with a partial pivoting LU
decomposition at the recommendation of section 2.4 of the paper.
More tests and a separate JNI function for each DARE solver variant were
added.
This method is used to check if the given value matches an expected value within a certain tolerance.
Co-authored-by: Tyler Veness <calcmogul@gmail.com>
Co-authored-by: Ryan Blue <ryanzblue@gmail.com>
Adds a reset method where teams can pass in module headings for the kinematics object to use if it gets an all-zero ChassisSpeeds while converting ChassisSpeeds to module states. Also removes internal states array, replacing it with an internal headings array.
I timed the DARE unit tests, and the new solver is 0 to 100% faster in
all cases (that is, it's at least as fast as Drake's and up to 2x faster
in some cases).
The new solver is also much simpler, takes less time to compile, and
drops the libwpimath.so size from 325 MB to 301 MB.
I think most of the compilation time is coming from the eigenvalue
decompositions used to enforce argument preconditions.
RKDP is strictly better in terms of accuracy per unit of work. We used
RKF45 for sim physics in the 2021 season, but we transitioned to RKDP
before the 2022 season.
Reverts "[wpimath] Constrain Rotation2d range to -pi to pi (#4611)"
This reverts commit d1d458db2b.
This broke multiple teams code in beta. It is also easier to limit the angle externally, then reconstruct a larger angle that got limited. This additionally adds comments to clarify the behavior and retains tests that were added in the reverted commit, and fixes a javadoc comment angle reference.
The ComputerVisionUtil class was added before AprilTag support was
announced. Now that it has, the functions for estimating a pose from
range and yaw are no longer needed; it's just better to get the pose
directly from the AprilTag.
The coordinate system on some function arguments was confusing or didn't
match the NWU convention the rest of the library uses. It's easier to
remove the functions now and add them back after they're fixed since the
fixes aren't trivial.
The range function was removed because it uses pitch and yaw in the
camera's spherical coordinate system, which is obsoleted by AprilTags.
AprilTags give you a 6DOF pose directly, so range can be obtained via
Pose2d.getTranslation().getDistance().
Fixes#4757.
